A couple days ago our minds were blown by a diminished reality demonstration showing objects being removed from live video feeds. Today’s mind-blowing video is a demonstration of MovieReshape, an image manipulation program by German researchers that’s going to make it much harder to believe anything our eyes see in future videos. As you can see in the demonstration above, the software allows physical characteristics of a person in a video to be manipulated by simply dragging sliders around.

It’s a pretty interesting — albeit scary — glimpse at where technology is headed.

(via f stoppers)

HBO posted this interesting behind-the-scenes video that gives a glimpse into the kind of special effects that went into filming the popular miniseries John Adams. It’s pretty crazy how they construct entire realities around the actors using CGI.

And you thought Photoshopping was hard…

We’ve all heard of trickery that goes into many kinds of photography to make something look more appealing to consumers, whether it’s a Big Mac at McDonalds or the swimming pool at a motel. Sometimes the discrepancy isn’t worth complaining about, but this wasn’t the case for David Ng (currently a guest blogger over at Boing Boing) when he purchased the Banzai Wild Waves Water Park. He took a photograph of the box art and then a photograph of his two children standing next to the actual inflatable water slide.

What we learn is that the product photographer used tiny children when photographing the slide. Just kidding. Unless the photog actually hired six miniature-yet-perfectly-proportional children as models, this is a pretty nasty case of dishonest photo-manipulation.

Reviewers on Amazon agree, giving the product two out of five stars. Here’s a sample comment:

It is a tiny piece of junk. I know a lot of things are mildly digitally enhanced these days but I have studied the picture and there is just NO way that is the product. I mean I dont know anything about doctoring photos but it is so grossly obvious in this case even a kid could tell. My 3 yr old can barely slide down b/c the other end of the pool is in the way.

At least the box says “product may not be as appears on image”.

(via Boing Boing)

Peter Funch is a New York City-based photographer who we featured a while back in a post titled “4 Creative Projects that Bend the Reality of Street Scenes“. Funch photographs scenes for extended periods of time, and then combines people who share something in common. In the photograph above, he chose to include only people who were carrying manila envelopes. You can see more of his work over at his V1 gallery page.

I recently came across this interesting interview video that gives a neat look at how Funch works and how the images are assembled. The interview itself is in English even though the introduction is not:

We learn that Funch shoots at street corners for 10-15 days at a time, and sometimes plants his tripod in the middle of the street with cars behind him. Interesting.

Image credit: Photograph by Peter Funch

Last week when Reuters released photographs of the volcanic activity at Eyjafjallajökull in Iceland, one photograph stood out to Wade Laube, the photo editor of The Sydney Morning Herald.

After making a few calls, Reuters decided to investigate. Laube writes on his blog,

Reuters had made contact with the photographer, an Icelandic local, and sought access to the original. It transpired that before being acquired by the wire service, the photograph had been in the possession of an Icelandic newspaper and it was there that some fairly liberal digital dodging and burning took place. When a comparison was made with the original, it became obvious that post production had been applied to sufficient extent that it violated Reuters’ very firm position on digital enhancement. So they retracted the picture and supplied the original in its place, and we dropped that image into the Herald for later editions.

Looking at the before and after photographs shown above, you can see that post-processing was done in order to make the plume of ash look extremely dramatic.

What are your thoughts on how far post-processing can go before it becomes too much?

Image credits: Photographs by Reuters

A banner on Andrew Romanoff’s Senate campaign website provoked a bit of an outrage from a minority group when people realized that it was digitally manipulated.

The original photo shows Romanoff, the Democratic candidate from Colorado, at a campaign kickoff.

The photo for the banner was tinted blue, and then had several people from other images spliced into it, presumably in order to make Romanoff look like he was surrounded by more followers. Some of the added people include an African American woman to the right of Romanoff, as well as a Latino American man towards the center of the banner. The photo changes caused some minorities to believe Romanoff was trying to appear like his supporters were more diverse.

The Photoshopped African American woman, Andrea Mosby, told reporters that she has no problem with the photo-tweak, since she supports Romanoff and was at the same rally.

Others disagree on principle. According to the Denver Post, Colorado minority leaders wrote to Romanoff’s campaign, expressing that they were “shocked, disturbed and outraged” that the campaign felt the need to manipulate an image to appear like he had more minority supporters. Some 25 people signed the letter that called the candidate’s integrity into question, saying:

“We are NOT random people to be moved around for aesthetic reasons…We are NOT political pawns to be used when convenient nor do we accept being manipulated and repositioned when it serves one’s political motives…The Photoshopping in of minorities is not acceptable and falls far short of the integrity we expect of candidates running for the US Senate.”

Romanoff’s campaign said that the banner was intended as a photo collage or montage of the event, designed by a volunteer to reflect the overall attendance at the campaign event.

Romanoff later removed the image from the site and issued an apology:

This decision and a description of it have caused offense. I regret that and have removed the montage from our website. I take offense at any suggestion that our campaign attempted to deceive anyone. That’s outrageous and false. I bring a lifetime of commitment to equality and opportunity, and I reject these attacks on my character. I am very proud of the diverse support we’ve already received and continue to earn every day.

(via Denver Post)

World Press Photo has disqualified one of the winners of this year’s contest after concluding that the photographer digitally manipulated his work. The disqualified entry “Street fighting, Kiev, Ukraine”, shot by Stepan Rudik for the Russian news agency RIA Novosti, had won 3rd prize in Sports Features.

This year, for the first time, photographers were required to submit RAW image files if the judges suspected that photographs were manipulated beyond what the rules allowed. The rule states:

The content of the image must not be altered. Only retouching which conforms to the currently accepted standards in the industry is allowed.

According to the British Journal of Photography, the manipulation involved removing the foot of one of the subjects in a photo.

(via Amateur Photographer)

Update: Stepan Rudik just contacted us with the original photograph and the version he entered in the contest. He says,

Your website gave information about disqualification of my material at the World Press Photo contest. I do NOT argue the decision of the jury.

At the same time, I would like to present the original photograph, from which it is clear that I haven’t made any significant alternation nor removed any important informative detail. The photograph I submitted to the contest is a crop, and the retouched detail is the foot of a man which appears on the original photograph, but who is not a subject of the image submitted to the contest. I believe this explanation is important for my reputation and good name as a reportage photographer. I’d like this picture to be published.

Here is the photograph in question:

It was actually a crop of the following photograph:

It wasn’t the crop, nor the post-processing, that caused the photograph to be disqualified, but the removal of the portion of the foot that is visible between the thumb and fingers of the hand being bandaged. We’ve cropped it ourselves here (Hover your mouse over the image to compare it to the version Rudik submitted):

Do you think the disqualification was justified? Share your thoughts with us in the comments!

Image credit: Photograph by Stepan Rudik

If you’ve ever wondered just how much editing goes into a particular photograph, there’s now a super easy way to find out for yourself. Image Error Level Analyser is a simple web application that takes a URL to a JPEG photo and returns an image showing differing “error levels” in the image. Here’s an example they give:

After submitting the image on the left, they return the one on the right.

Error level analysis shows differing error levels throughout this image, strongly suggesting some form of digital manipulation. Areas to note are the lips and shirt, as well as the eyes. All are at significantly different error levels than their surroundings. Presumably, colours have been altered and areas brightened.

Thus, you can now investigate any JPEG you find on the Internet to see roughly how much the photograph has been edited or manipulated. The app even gives you a permalink to the resulting image comparison. Try it out, and post your permalinks in the comments to share your findings with us!

Update: Here’s an interesting article by Wired on how researchers use this technique.

Here is a quick look at some interesting photography projects that present a manipulated view of reality, along with thoughts on how they were done and how you can accomplish the same thing.

#1: empty L.A. by Matt Logue

This project, created over a four year period starting in 2005, gives an interesting glimpse at what Los Angeles would look like if people and their vehicles instantly disappeared off the face of their earth.

There’s are a few ways you could accomplish this.

First, you can stack neutral density filters on your lens to enable extremely long exposure times. The long exposure would cause everything that moves (i.e. people and cars) to disappear from the scene.

However, it doesn’t look like Logue employed this technique, since the clouds in his photographs are clear and well defined. You would also expect trees in long exposure photographs to be soft and blurred, since the leaves are constantly moving.

A second option is to take a very large number of photographs, and then use an image editor to combine only the portions that don’t contain any people or cars. Doing this at a time when the road is least busy would obviously be easiest if editing by hand, though Logue has quite a few shots from busy hours of the day. Photo editors like Photoshop or Enfuse can also help you automatically stack images and filter out non-constants.

Visit the project page to see the rest of the photographs and/or to buy the book.

#2: Tokyo Nobody by Masataka Nakano

Japanese photographer Masataka Nakano spent 11 years shooting photographs of Tokyo devoid of people.

This project is unlike the other three in that no clever image manipulation was done. Nakano visited ordinarily busy locations during times of low activity (i.e. major holidays), and patiently waited for just the right time to make each of his photographs.

You can also purchase his paperback book on Amazon.

#3: Babel Tales by Peter Funch

At first, many of Peter Funch’s New York City street scenes seem ordinary. Then, as you look closer, you begin to realize that in each one, there’s something eerily similar with everyone in the scene. Perhaps everyone is holding a manila envelope, or is smoking, or is running, or is looking in the same direction, or is yawning, or is homeless.

You get the idea.

The amazing thing is, these photographs weren’t staged, but rather manipulated.

Funch took each of these photographs two weeks at a time, found people in the frames who had something in common, and stitched the photographs together using an image editor.

While the first two steps (shooting and selecting) are time consuming, the third step (stitching) is what’s difficult. The idea is similar to what we outlined in our “cloning yourself” tutorial, but rather than having a static background that you can easily mask over, stitching this type of photograph might require a pixel by pixel degree of care, since people are constantly walking around in the frame.

You can tell that this stitching is the technique used by Funch by observing that in some of the photographs, the lighting is different for various people in the frame.

It’s somewhat mind-boggling compared to the first two, but still doable.

Check out the rest of the photographs in the series in this gallery.

#4: Orderly Conduct by Bahbak Hashemi-Nezhad

Bahbak Hashemi-Nezhad shot his orderly conduct series of photographs in three locations: London, Amsterdam, and Toyko. The idea is identical to what Funch did in NYC.

Most of Hashemi-Nezhad’s photographs are easier to accomplish than Funch’s for a couple of reasons.

First, many of them are taken indoors, where the lighting conditions don’t change. This makes it much harder to detect image editing, since there aren’t lighting difference between people in the indoor scenes.

The sparseness of many of the scenes also makes for easier editing. Many of Funch’s photographs were taken up close with people moving around in the frame and overlapping one another. Many of Hashemi-Nezhad’s photographs have minimal overlap between people, and constant backgrounds. This makes it much easier and less time consuming to stitch the photographs together.

For the rest of the Orderly Conduct photographs, you visit his website here.

Hopefully you saw something new in this post that you hadn’t seen before, and that it inspired you to experiment with some new ideas.

If you have any comments on this kind of project, please leave a comment! We’d also love to hear from you if you have links to other similar projects!